Transcript for "Demystifying AI: Workday's Innovations in Action": Good afternoon, everyone, and very welcome to our session on demystifying AI this afternoon. When I mention artificial intelligence, what comes to mind? A distant futuristic robot? Perhaps a cutting edge chatbot? The truth is AI is here, and it is the single most transformative force operating in your life right now. From the moment you woke up this morning to the way your business is run. Over the next hour or so, we're going to demystify this ubiquitous technology and, hopefully, get you to see it as not just a complex algorithm, but as a powerful partner to you and to your business. My quick way of introduction, my name is Jannie Malan. I'm a solution manager with Workday. I've been with Workday for, seven and a half years, and I work out at the Johannesburg office in sunny South Africa, although it's quite rainy today. Workday also afforded me the opportunity a year or so ago to complete the Oxford artificial intelligence program, for which I'm very, very thankful. And, hopefully, you can impart some of that knowledge to you today as well. So let's quickly look at our agenda for the, for the session. First of all, we just wanna get a sense of where we are currently in terms of the return on investment for AI. Then we wanna look at some basic concepts that prove to be useful when we talk about artificial intelligence. I'm also going to share with you Workday's approach to artificial intelligence. We're going to unpack the agent system of record, the way to manage all of the agents that you're likely to have in your in your organization moving forward. And then we're gonna talk a little bit around next generation ERP, the road that Workday is heading on. Please also note that I might make some forward looking statements during the course of this presentation. Now I have no doubt that many of you, like myself, use AI during the course of my day. These days, AI is seamlessly woven into the fabric of our daily routines, often without us even being consciously aware of it. When you unlock your phone with facial recognition, you ask GPT a question, or you see personalized recommendations on streaming services and online shopping sites, you are interacting with AI. From predictive text on your keyboard to smart algorithms curating your social media feeds, AI is constantly working behind the scenes to make everyday tasks more efficient, more personalized, and ideally, more convenient as well. From a study of thousand global business decision makers telling us what leaders are thinking, feeling, and doing when it comes to AI, we recognize that almost 80% of enterprises believe that AI has or will soon disrupt their business. 73% feel pressure to increase the amount of adoption investment that they're currently putting into AI. And then finally, and probably most telling, is that almost 3% of business leaders are in the process of implementing or exploring AI within their own organizations today. Of course, the use cases that they're realizing are very, very diverse. And many organizations have actively started applying, AI capability within their own organizations. In fact, reports by analysts like McKinsey clearly indicate that businesses intend to increase their investment in agentic AI significantly over the next three to five years by as much as 93%. However, as we can see from headlines like the ones on screen, many organizations are not yet realizing the projected ROI from their AI initiatives. Reports assert that a stunning 95% of these pilot projects are failing to deliver any measurable return on investment. And only a small fraction, about 5% of companies, have managed to achieve a meaningful, measurable impact on their artificial intelligence investments. Now the Gartner Gartner hype cycle for AI goes a little way to explain the maturing landscape around AI technology. While still a transformative force, generative AI has created has kind of crested, moved over the peak of inflated expectation. We had high hopes for it. And it's descending into what we call the trough of disillusionment. Now this signifies a period where the initial hype is being tempered by the practical challenges of implementation, cost, and of demonstrating a clear return on investment. The emphasis is now shifting towards identifying robust use cases and realizing tangible business value from generative AI projects, often in combination with other AI techniques. Organizations are also recognizing that a singular focus on Gen AI is not sufficient. The hype cycle underscores the growing importance of composite AI, which binds various AI techniques, for example, machine learning, knowledge graphs, natural language processing, and the like, to go ahead and solve more complex business problems. And this is a normal evolution because some people think AI is just a technology that simply needs to be implemented. Unfortunately, that's not quite the case. AI is truly industrial purpose technology. And I'll I'll talk about general purpose technology just now. But it means that it is affecting many, many different fields. It's changing the needs of customer, your industry, the expectations of your employees, and your organization. To be part of that successful 5%, you need to start with some basics. First of all, do you have the trusted, consolidated, and real time data that generative AI requires to function? If not, your AI will fail. In addition to that, do you have employees and managers that understand AI and its implications? And it has a skill set in both AI as well as the use case pertinent to your specific organization and industry. To set yourself up for success with AI, AI should ideally be in the natural flow of work. And this is very likely means that you may need to redesign and standardize your processes to achieve this. Now fortunately, with Workday, our clients can address all three of these. They have trusted data set. They've got real time visibility on their employee talent. And they have the ability to easily standardize and revise their processes to better meet the requirements of AI enablement. Let's take a quick look at some basic concepts around AI that can prove to be helpful to us. Now I mentioned it just now in general purpose technology, and let's quickly talk about what they call GPTs. Now GPTs are typically innovations that have a profound and broad impact across the entire economy and society. Unlike niche technology that improves a specific task, let's say, for example, a new drill bit in the mining industry, a GPT is a foundational building block that enables a cascade of further innovations. This technology is typically, technology is typically classified as a key if it meets the following key criteria. First of all, it can be used across many different sectors and industries. Secondly, it has the potential for continuous improvement over time, leading to lower costs, greater efficiency, and new applications. And then thirdly, it stimulates the creation of new complementary products, processes, and business model. Now throughout history, there have been quite a few technologies that have fundamentally reshaped the world by functioning as GPTs. And I'll go all the way back to the steam engine, which transformed manufacturing, transportation, and agriculture, leading to the industrial revolution. And then we've got electricity that revolutionized factories by allowing for flexible production lines, and powered a wave of new inventions from the telegraph right up to household appliances. And then, of course, the Internet. It connected the world, creating entirely new industries like commerce and social media, and drastically changed the way in how we communicate, how we work, and how we access information. Now artificial intelligence, particularly generative AI, is widely considered the latest and perhaps the most important GPT. It has the same transformative potential because it isn't just one product, but a set of versatile tools that can be applied to almost any task involving data and logic. And to that point, a quick example of myself of mine close to home. Over the past few years, we've had great challenges around the poaching of rhinos almost to the point of near extinction. Over the past few years, we started embedding AI technology into things like AI enabled collars, AI drones, and also using AI to look at the datasets of migration patterns, etcetera. And fortunate enough, this has meant that we've been able to reduce the numbers of poaching on a year on year basis. Just to give you a little bit of an example of how a GP team can apply to different technologies and to different industries. But let's also just have a quick working definition around, artificial intelligence. And this one is brought to us by artificial intelligence itself. This was given to me by Jannie. But really, artificial intelligence, we're talking about, the ability for a computer to perform tasks that would typically require human intelligence, about giving machines the capacity to think, learn, and act in ways that mimic human cognitive abilities. Let's also touch on how we got here with AI. Now don't be worried. I'm not going to take you through the entire journey and history of AI. But what I do want to suggest here is that you use your favorite AI assistant. In my case, this might be notebook l m, and ask it to create a nice podcast for you about the history of AI that you can then easily consume while driving or while you, are doing some other task. The journey of AI can, at a high level, be understood through some distinct wave tides, as you can see here as they've been broken down, stages of development. And these were often influenced by the AI hype cycle that we saw earlier, but also punctuated by periods of reduced funding and interests. And these are known as the AI winters, in which there was very, very little development around AI. The important thing here is, and what should be clear is that AI is by no means a new technology. So if AI has come such a long road, why have we seen it come to the forefront so rapidly in recent years? Well, there's a few key factors that contributed to the rise of AI as we see it today. The first one being data. The modern world is generating an unprecedented amount of data. Every online transaction, every social media post, every connected device contributes to that low ocean of information. This big data is the fuel for AI. Machine learning models need vast datasets to learn from. The more data they're trained on, the more accurately they can identify patterns, make predictions, and perform complex tasks. The next factor is hardware. Processing these immense amounts of data, as you can imagine, needed for AI training, requires some really, really powerful hardware. Recent advancements in computer hardware, like the development of graphic and tensor processing, have provided this crucial component. The rise of cloud computing services has also made this power accessible to researchers and companies without the need for massive costly on-site infrastructure. So that's kind of democratized AI development, allowing more people to experiment and to innovate in this space. And then there's also been a shift from isolated words to contextual meaning. Before recent advancements, AI models often use what we call a bag of words approach. This meant they saw each word in a sentence as a standalone unit, failing to recognize how surrounding words changed the meaning. This led to frequent misunderstandings, especially when we were using ambiguous words. The key breakthrough came with the development of the transformer architecture and self attention mechanisms. These new models, such as BERT or GPT, and I'm sure you'll recognize that, were designed to process entire sequences of words at once, allowing them the ability to understand context. Now the superpower of AI is not logical thinking, but it's recognizing patterns in large amounts of data. It links these patterns to make predictions or to create new content. So what do I mean by that? Well, let's think of some of the patterns that we as humans can recognize on a day to day basis. We can recognize patterns and shapes. So we might ask ourselves, you know, for example, what is the next shape? In this case, yes. It'll be a, triangle. We can also recognize, patterns in data. So in this case, recognizing the real estate prices, and projecting that into the future. And then also recognizing patterns in text. So determining statistically the probability of words appearing in relationship to each other. So in a way, AI is like a fortune teller who doesn't really see into the future, but knows the probabilities of the past really, really well, and can subsequently make those informed predictions. So we understand that AI is really good at discerning patterns. But how does an AI learn these patterns? Is it always the same learning processes, or do they differ? And all I'm gonna do is to quickly unpack three of the main ways in which the AI learns these patterns. And let's start off by looking at what is known as supervised learning. As the name suggests, the AI learns exactly in the same way like a student might. It receives examples and the correct answer. A classic example that we all know is span detection in our inboxes. Through classification, the AI learns that an containing words like gambling or sweepstakes are usually spam. Another example of this, application of this is regression. Here, the AI learns to predict a value. An example outside of the HR sphere is price action for stocks or real estate. Within HR, we can use this to predict staffing. Then there is unsupervised learning. Here, the AI has no predefined answers. It simply receives data and must independently find patterns and structures. The example of customer groups illustrates this technique. The AI identifies similar behavior and groups customers into segments by itself. For HR, this is extremely relevant for identifying employee groups with similar career paths or similar training needs. And that then helps us to create personalized development plans. And finally, reinforcement learning. You can think of this almost like a reward system. The AI learns through trial and error, as many of us learn as well. It receives a reward for good decisions and a punishment for bad ones. The most famous example of this is AlphaGo, which is an AI that learned the complex game of Go without any human guidance simply by playing millions of games against itself. So let's take one of these methods. We'll take the most widespread method, supervised learning, and we'll quickly look at it in a little bit more detail. It's the cornerstone for most AI applications we deal with today. Think of applicant selection, for example, or someone applying for a position with an organization. We'd use that as an example, and we'll go through the process step by step. In the same way that you might bake a cake, there's a recipe. And like a recipe, the AI model also consists of two main ingredients. First, the AI needs training. These are this is historical data for which we already know the correct answer. For example, we've got hundreds of resumes that we have already marked as suitable or not suitable. This data is then processed with a suitable algorithm. The algorithm really is the brain of the model, the actual learning instruction. The result of this equation is our AI model, a finished trained system that has learned to recognize certain patterns. But that's only half the battle. When we finish the model, it is now unleashed on live data. So, for example, new or unseen data. In our example, this would be the new application that has just come in. The model applies its learned knowledge and makes a prediction. It tells us this candidate has a high probability of being a good fit for the role. And here, an absolute crucial point comes into play, the human control or the human influence. The AI merely provides us with a suggestion, a well founded forecast. The final decision is always made by you. We check the prediction and decide whether to follow the suggestion. This then leads to an outcome. For example, a invitation to an interview. And year the circle closes. The outcome, for example, whether our decision was correct, can in turn be fed back into the process as new training data. In this way, the system continuously learns, and it gets better over time. Essentially, machine learning is about creating algorithms that can learn from and make inferences in data, enabling systems to adapt and to perform tasks really, really intelligently. But just some quick definitions here as well. You'll notice there the large language model or the LOM. Now a large language model is a computer program that can fit enough exam examples to be able to recognize and interpret human language and other types of complex data. So you will will likely be familiar with ChatGPT, Gemini, Grok, Claude, just to name a few LLMs. Now many LLMs are trained on data that has been gathered from the Internet. Thousands or millions of gigabyte worth of text. I also want to, again, point out the importance of context when it comes to AI. Now a generative language model's output is only as good as the combination of data and context. So the data, what the AI has learned, the patterns, okay, and the context, specific situation or the exact instruction that you are giving it. So imagine asking a large language model about the apple. Okay? Without context, it's going to potentially refer to the fruit. But if you provide context, say, for example, say I'm a software developer, the LRA will understand you more likely referring to Apple software or products. And, of course, you can add additional context. This shows the relevant context significantly of the accuracy and usefulness of your large language model responses. And to that point, I would urge you to introduce yourself to your large language model that you prefer using. Give it the context about the job that you do, the results that you prefer, and that's really going to improve, potentially, the results that you get from your LLM. Excellent. So we've looked at some of the foundational components, pertaining to AI. And, hopefully, we've got a little bit of an understanding of of how AIs learn, how they recognize patterns, and how can they can potentially be applied within our, private lives and also within our businesses. What I'd like to do now is to unpack Illuminate. Now Illuminate is what we call our Workday AI offering. Illuminate unlocks the power of an organization's people and data, accelerating manual tasks, assisting every employee, and ultimately, transforming the entire business process. And there are three main ways in which AI and Workday is experienced. And we call these accelerate, assist, and transform. And I'm going to look at each of these in a little bit more detail, so that you can kinda see our thinking in terms of where we wanna use AI to enable and assist employees to do their work better and more effectively. Of course, one thing I do wanna point out, the importance here and the benefit that the Workday clients have, is that they're transacting on that single system of record. The single system of record for their people, their money, and their agents. And that brings that vital context to the LLM because you're transacting on the data that is pertinent to your organization and not simply on the data that is pertinent, you know, right across the Internet and is extremely generic. Right. So let's have a quick look at Illuminate. Now we said the first thing that Illuminate does is, to accelerate. Basically, getting people to do the work that they do on a daily basis faster. Now, accelerate refers to using technologies such as generative AI to get work done quicker, saving you time. And, I mean, saving time equates to money. As an example, instead of taking several hours to create a job description, you can now go and do it in minutes. Accelerate enables workers to do their existing work more efficiently through capabilities like automation, or content creation, or summarization. I'm sure many of you today are using, the summarization capabilities that you have in your online meeting, software to summarize those meetings for you instead of taking those notes yourself. AI is used to streamline those mundane pros, who are tops, enabling workers to use their skills on more strategic and more creative tasks that require human input. And also, just give people more work satisfaction as well. Now, the example that you see playing here on the screen, is an example of Gen AI being used to create a job description. Taking a time consuming task, reaching the rich textual data we have in Workday, and the understanding of your business, your language, and how you structure your job descriptions to create a robust draft in seconds, that'll be a great starting point. You can obviously then manipulate it, make it more informal or more formal, and slant it even more to your organizational requirements. But at least you're not starting from a blank slate. Of course, you could potentially use ChatGPT or another agent to do this, but the job description won't reflect your organization, your language, your tone of voice, etcetera. Let's look at assist. Now assist reduces the cognitive load on workers by providing conversational search, Not only finding what they need, but guiding during task completion. Now this is really like having that assistant next to you. You know, when you're confused, when you don't know how to do a certain task, being able to reach out and get assistance from the Workday assistance. There's no there's so many things today, so much data that managers, employees need to be aware of, that it becomes really, really difficult to track of everything. The Workday assistant also recommends actions that provides context for the task at hand. So it's really within the flow of work, helping organizations with more accurate and more consistent decision making. Again, it's using natural language. That means you don't have to remember what a task is called in a system or how to complete the steps. It fits naturally into the flow of work with the assistant identifying your intentions and walking you through the task, even prepopulating fields using your current context. Now the screens depicted here is just an example of showing the Workday assistant that is currently helping to guide a manager through the process of making a one time bonus payment to one of their direct reports. The assistant can suggest values to prefill forms and explain why it's prefilling a specific value. The assistant can also use data and access to policy information to the manager know the allowable range for a bonus payment for this specific employee. A dramatically different and much better experience for a manager rather than having to go around and and look for information all over the place. So we've looked at accelerate. We've looked at assist. And now we can look at transform. Now transform is where AI can lead to exponential improvements by using AI to completely rethink how work gets done within your organization. Agents, agentic, AI falls into this category where work gets done on your behalf. The truly transformative impact is felt when agents all connect and collaborate with other agents to perform complete tasks to meet your overarching goals. Now they I'm gonna have more on agentic AI and the agent system in the record shortly. But this is the three ways in our approach, our Workday approach, in which we look to bring the benefit of artificial intelligence into the platform, and also to ensure that we're getting that return on investment for our clients. At Workday, we recognize the challenges of AI adoption. Success with Workday Illuminate means adhering to five firmly rooted technological and ethical principles. First of all, keeping the human in the loop. Workday's AI is designed to empower humans, not replace them. Collaboration between humans and AI recognizes that each has unique strengths and limitations. Being value centric ensures that AI the AI that we deliver provides real value. We work directly with our customers to define and measure the impact of our AI innovations through brainstorms, advisory councils, early access programs, and design partner groups. We know, of course, that high quality data is essential for AI, and it is critical to assist confidence with end users. We understand that quality, consistency, and accuracy are equally important as measures of success. We've extended and, in fact, opened up, as you will see later, our platform to enable customers and partners to build custom AI applications. Our AI is developed responsibly by design throughout the entire software development life cycle. In addition, as organizations increasingly embrace AI, they need to adopt artificial intelligence frameworks to govern usage. For this reason, Workday has adopted four core principles for responsible AI, with the aim being to ensure our AI solutions are ethical, fair, and trustworthy. And these principles are to amplify human potential, to deliver on our commitment to data privacy and protection, to positively impact society, and to always, through AI, champion transparency and fairness. So you've gotten a bit of a glimpse of AI illuminate, how will they approach to AI? And now I'd like to move to this exciting theme of the agent system of record. I think we're at a very interesting mission, and business and people leaders need to recognize that this is the last generation that will manage only human resources. Further, future generations will manage both human and AI resources. Potentially, hundreds of agents across multiple functions. We often refer to this as, agent sprawl. Now agents, or digital employees, if you prefer, will need to be managed in similar ways to how we currently manage our human employees. When you look at the key capabilities needed for employees and for agents, whether they're human or artificial, the needs are actually surprisingly similar. First of all, there is life cycle management. You need centralized roles. You need skills. You need onboarding and offboarding. You have to track and measure the goals and performance of those employees, whether they're digital, digital, or human. There is workforce optimization implications. You still need to identify, staff, and orchestrate talent to achieve priorities. And, of course, there's always governance and compliance. Compliance with laws, access controls, ethics, and relevancy. Now fortunately, because we already have extensive experience in managing your human workforce, Workday is ideally positioned to be your one stop of record to manage both the human and the digital workforce. It will give you one place from where you can onboard all of your AI agents, assign roles, manage their role based tests, and tune their skills. So, again, just to quickly touch on agentic AI, it refers to the artificial intelligence systems designed to act autonomously and proactively to achieve specific goals that we require them to do. Task based agents are designed to excel at automating repetitive or complex tasks, like processing invoices, freeing up for human workers for other activities. Role based agents are designed to support humans by understanding the role complexities and level of access within an organization by taking on very specific tasks and responsibilities. Now we're working on multiple agents within Workday. We're always keeping a human at the center. Always using AI to amplify human potential. Now we believe that we are approaching agents in the right way because, ultimately, all this comes down to one simple thing, realization of value. Every decision in how we approach agents is centered around customer value. Now we have several agents in general availability today, and some of the ones that you see here pertaining to contracts. So the ability to, apply or add intelligence to contracts, which is usually static documents. We take that rich data within your documents, and we apply artificial intelligence to it. And also, agents that apply to talent mobility. And there are more agents being released every single day. The next wave of agents span the full range of next generation ERP, covering everything from running a financial close, to running payroll, to doing performance reviews. And we're gonna look at a few examples quickly as well. One of the first ones I wanna quickly show you is the frontline agent that can provide an automated experience across time, absence, and scheduling for frontline workers and managers. When you might have often changes in schedule, it can be quite a cumbersome experience. Automation here can cut process times from fifteen minutes right down to two. Let me just quickly click ahead here and get this playing for us. So what we're seeing here is the ability for an employee to indicate that they cannot, attend a specific shift. The agent can automatically pick that up, automatically route that to the relevant shift manager or team leader, automatically suggest, other employees that might be able to fill in and step into that position, automatically reach out to those specific employees to find out whether they will be able to attend that specific shift, get them enrolled into that shift, and basically ensure that that shift is fulfilled, with no human intervention aside from the fact that there needs to be approval, from the shift leader or the manager's perspective. The next one I want to quickly show you an example of is our financial audit agent. So really, in this case, facilitating communication with auditors, adding control and monitoring, and adding risk creation when it comes to financial. But I'll have my colleague quickly explain this for us. As the system of record, we expect to have much of the data and now documents needed to respond to audit requests. But there is still a lot of manual work for accountants as they prepare to respond to their auditors. Based on our research, we believe an audit and compliance agentic AI solution can help transform this and is informing our current thinking as we develop a future financial audit agent, which will help automate evidence collection and even support communication with auditors. Imagine the burden lifted from accounting teams. Let's take a look. As the controller, I'm starting with our future audit hub, where I see the progress our financial audit agent is making in response to audit requests as well as the agent's effectiveness in saving the team time. Additionally, the agent is identifying risk in our internal controls and any challenges with our ledger integrity. Additionally, I see supporting documents provided by document intelligence like the 10 k, board meeting notes, and more. Our agent alerts us to a new audit request from the auditor. They've shared a spreadsheet that contains a list of their required documents. This is where our financial audit agent gets to work. It transforms the spreadsheet into an audit request work area and assigns responsible preparers and reviewers. In some cases, there is no judgment needed, so the agent can perform both roles, preparer and reviewer. In other areas, there is a greater degree of judgment where the accountant needs to review. The agent has gathered documents and reports, mapped them to the requests, and updated the status, notifying the accountant to verify the details. If we go ahead and verify the agent's work, we see the documents sourced by the agent. If needed, we could add a comment, add additional documents. Once the review is complete, we confirm the readiness of this response and approve for submission back to the auditor. The result, with an agent driving audit evidence, the time needed to respond to an auditor to prove the integrity of controls and financial statements will be dramatically reduced. Excellent. And then I want to look at a third example. A Workday Optimist is a collection of AI powered capabilities and agents that support Workday admins in maintaining and continuously improving the value and experience of the Workday investment. So we're kind of turning the AI inwards to look at how we can improve the platform itself. So if we get this started, we can quickly talk through this. There's go. And we what we're looking at here is the ability to use the agent to improve our business processes. Because often, you might find that there might be some efficiencies, too many steps, people not approving, or whatever the case might be. In this example, I'm asking the agent to identify the top the the the worst performing processes within my Workday tenant. I can then ask the agent to compare it to benchmarking information. Remember, because we're one single system, we've got those benchmarks and understand where the optimized processes sit across our global customers. So I can quite easily compare to those optimized processes to identify any inefficiencies in my own business processes. As you see, I can then ask the agent to suggest to me, where those, those problematic areas are and to go even further by recommending ways in which I can optimize and improve my existing business processes. In doing this, ultimately, leading to improved efficiencies, quicker turnarounds, more standardization, and improvement of the investments that I've already made within Workday. Now when you refer to the Workday agent system of record, this is the centralized platform designed to manage, govern, and optimize an organization's entire fleet of AI agents. It acts as a single system for managing your digital workforce alongside your human employees. The platform allows businesses to onboard new AI agents, define their roles and responsibilities, track their impact and cost, and ensure compliance with a next generation open AI ERP platform. Workday's unique differentiator here is the context and the cleanliness of our data. So when you build an ecosystem, it has access to far more contextualized business rich data than others. At the heart of the Workday agents of record is a central place for agent registration. The command center with a clear view of every AI agent available to your teams in your organization. You can register agents, see the description and their availability. You can control access based on the agent's role and activate or deactivate them across the entire organization. You can register agents, see their description, their availability, and even control access based on the agent's role, just like you might do with an employee. Need to deactivate an agent because they're not performing? Simply remove it remove its access across the board and across the entire organization. And get insight into the performance of your agentic AI. There's something like this agentic analytics dashboard, which gives you a clear indication of performance levels, efficiency impact, and how well they're collaborating with your human employees. This, as you can imagine, really starts informing and starts driving the ability to get that return on investment on AI that we are all looking for. Now let's talk a little bit about about what we refer to as the next generation ERP. Using the data and context for how work gets done allows us to focus three key areas for evolving Workday as a platform. First, building the best agents for people and money. It is a process that deals with people or money in your organization. We want to provide the best solution, the best agent with the deep view for that specific problem. Secondly, we're opening up the Workday platform so you can build agents, partners can build agents, and you can leverage your Workday data. The only limit is your imagination, but it's all managed securely within Workday. And then finally, a platform is really nothing without the community that supports it. Without its ecosystem, partners, and developers that go and build on that platform and really show the art of the possible on top of this innovation. So when we talk about next generation ERP for the AI era, this is the embodiment of that vision. It's moving worry and the ERP category beyond an application, beyond a ring fenced the walled garden, to a real force multiplier for work for the next generation. Let's quickly talk about the tools. So what we have done is enabled what we call Workday Build. Okay? And as you can see, there's several pillars here speaking to the tools, partner ecosystem, and the community. Let's quickly look at those tools. You can leverage Workday build agents and consume partner build agents, but we also want to create or give you the ability to create your own custom agents to solve for those very specific organizational or industry use cases. Now one of the tools that we've acquired with this purpose is a tool called FlowWise. It's an enterprise grade, low code, visual builder to easily and quickly create those agents in a drag and drop fashion. We've added this capability to the core of Workday Build within the developer site. In addition to that, you may be familiar with the extend capability within Workday that allows you to extend the the functionality within Workday to, again, cater for those specific use cases that you might have. You can extend the business process framework, datasets, the UX, etcetera. FlowWise agents can also mount on the extend apps, and you can start imagining the power that we can create there. The second pillar of WorkdayBuild, is our marketplace, our partner ecosystem. Well, the benefit here is for our customers, they can quickly solve for industry, geographic, and category challenges. For our partners, this means these programs let them build and distribute solutions at significant scale. In simplest terms, the Workday marketplace is a website for Workday customers browse for solutions that will extend the functionality of the Workday core products. The solutions listed in the marketplace are built by Workday and by our partners, and span across different persona, industry, niche business needs. The goal is to evolve into a concept comprehensive hub offering multiple purchase paths, full functionality gaps, meet regional needs, and reduce cost. And as you can see, there's examples in that mod in, marketplace that pertain to applications, agents, templates, integrations, and services. And the final pillar within the Workday build is an active community where developers can get certified, where teams can get skilled as part of an ecosystem that grows the talent pool of external Workday developers, providing that fuel for accelerating innovation and, specifically, AI innovation at scale. The year of passively observing just observing AI is over. The challenge for all of us, regardless of our industry, is to transition from being users of AI to becoming engaged, ethical builders of the AI enabled future. Now if there's three things I'd like to take away from the session today, It's the following. AI return on investment can be realized if it's done right, and if the necessary foundation is in place. You've seen have gotten a glimpse of how Workday enables responsible case based AI, and is your go to agent system of record. And we've also seen next generation ERP is enabling is enabled by what we call workday build, allowing you to create and manage your agentic and, of course, your human workforce. Before concluding, just a quick reminder from my side for those of you still able to do so. We've got our rising event, rising EMEA, happening, from the eighteenth to the November 20 in Barcelona, Spain. If you can still make it, it will be great to see you there. There's a wealth of information. You can engage with Workday. You can also engage many, many of our clients. So if you are able to do so, please feel free to join us there. And I trust that this brief introduction to artificial intelligence and Workday's approach to helping businesses adopt and realize the value of AI and its multiple manifestations has been insightful. We're happy to share more detailed information with you or to set up a workshop to understand your specific artificial intelligence challenges and, more importantly, opportunities. So thank you so much for your participation, and happy to take some questions. And I I think start with some of the questions that have already come through. So let's just have a quick look. We'll just go down here for us. Just bear with me for a second. Yes. There's a question around trainings and resources. On a continuous basis, we are making those trainings and resources readily available. If you are an existing Workday client, of course, all of this will be available within the community. Again, just watch our website and, just, be sure that you check the community on a regular basis. There's a question around whether this presentation will be available for distribution. Yes. We will certainly share that with you afterwards. Very good question around agentic AI. Now, the AI capability, the Illuminate capability is part and parcel of the core of Workday. So that's a huge, huge benefit. It's not an additional capability that you need to additional SKU that you need to acquire. However, when it comes to the specific AI agents, like, say, for example, Eversource contract management, or perhaps candidate management, or, the exciting learning agents that are coming down the road, there is likely to be a separate cost if you wanted to activate that specific agent. Good. I see there's a question around the clear breakdown of when each of the higher capability AI features might be available. And that is part of our roadmap items because there's going to be some new and exciting announcements at the upcoming rising EMEA. But, also, the road maps are available within the community as well, where you can go have a look at and look at more detailed information pertaining to each of these agents. A question around the AI tools for learning. Again, safe harbor statement applies. As you may have seen, we are in the process of potentially acquiring a solution called Sana, which is going to give us incredible robust agentic AI capability focused specifically on learning, and on knowledge enablement. So please watch this space. It's one of the ones that I'm very, very excited about, having worked significantly within the learning space, and the LMS and learning experience platforms. Very excited about that. There's also see there's a question around the Gen AI job descriptions. I know many of you might have job catalogs. And, of course, we can use that as a dataset to inform the large language model as well. So, again, one can have discussions around this. Alright. I apologize if I have perhaps missed any of the questions, but I just wanted to thank you again for your participation this afternoon. And if there are any more questions, please feel free to add them. Thank you so much for your kind remarks coming through. Really appreciate the feedback. Excellent. Well, we're coming up to time. Again, I just wanted to thank you very, very much for your participation, for attending our session, and for your input. And I wish all of you an amazing afternoon. All the best. Thank you.